IPNA clinical practice recommendations for the diagnosis and management of children with steroid-sensitive nephrotic syndrome

The definitions presented in this CPR agree with previously published IPNA Clinical Practice Recommendations for the diagnosis and management of children with steroid-resistant nephrotic syndrome (SRNS) [19] and the KDIGO 2021 Guideline for the Management of Glomerular Diseases [15, 20]. In addition, new definitions for treatment outcomes to help guide change of therapy, e.g., the introduction of steroid-sparing agents, are provided. Of note, patients with late response, i.e., remission between 4 and 6 weeks of PDN therapy, are defined as “SSNS late responder” and should be managed as SSNS but anticipating a potentially more severe course.

The proposed definition of frequently relapsing nephrotic syndrome (FRNS) differs from previous ones including those from KDIGO. The prescription for the first episode of SSNS usually amounts to a PDN exposure of ~ 115 mg/kg. Each relapse adds ~ 40–45 mg/kg; three relapses would mean 120–130 mg/kg, and four relapses would mean 160 mg/kg over 12 months. A child with 4 relapses in a year would thus be exposed to ~ 0.5 mg/kg/day PDN, which may not be acceptable in terms of toxicity risk. Therefore, we propose to revise the definition of FRNS to include children with 2 or more relapses in the first 6 months of the disease, or 3 or more relapses in any 12-month period. The definition of FRNS as a disease classification serves as a clinical indicator that treatment strategies should be transitioned from responsive, ad hoc therapy to preventive or proactive therapy to reduce relapses and corticosteroid toxicity. Considering the spectrum of steroid-associated adverse effects, the anxiety that the fear of relapses causes in patients and families and the patient/family preferences for steroid minimization, the rationale for this change is two-fold. First, the new definition of FRNS promotes a discussion and selection of therapy for patients with FRNS, which incorporates patient/family preferences. Second, the new definition acknowledges the fact that many pediatric nephrology centers throughout the globe already implement this threshold in routine clinical practice to optimize steroid minimization.

Regarding steroid-dependent nephrotic syndrome (SDNS), the wording of the definition has been fine-tuned. The term “recommended PDN” has been added to promote a uniform steroid treatment in all children with NS both in relapse and in remission. Moreover, “PDN for first presentation or relapse” aims to clarify that patients relapsing during or 14 days after low-dose maintenance treatment with PDN are not steroid-dependent. It is only a relapse during or within 14 days after completing high-dose PDN (i.e., 2 mg/kg per day or 1.5 mg/kg on alternate days) discontinuation, that qualifies for this definition.

Regarding the definition of hypoalbuminemia, usually a cut-off of 30 g/L is used. However, there is significant variation between serum albumin assays in different laboratories. The 2021 KDIGO guideline states: “Laboratory-specific values: serum albumin should be measured by bromocresol purple (BCP; colorimetric) capillary electrophoresis (CE), or immunonephelometric (iMN) methods. Bromocresol green (BCG) methods can give erroneously high results” [20]. The values of serum albumin measured by BCG are about 5.5 g/L higher than those measured by the BCP, CE, or iMN methods [21], so the definition of the degree of hypoalbuminemia required to meet a definition of NS varies according to the method used for quantifying serum albumin concentration. The bias between different albumin assays may affect clinical decision-making [22]. However, as long as a specific method is used consistently based on local laboratory practice, changes in serial albumin concentration can be monitored over time.

Regarding statural growth, we suggest using the definition for impaired statural growth as recommended for children with CKD, i.e., a height velocity < 25th percentile and/or height < 3rd percentile [23]. Height velocity should be calculated based on an observation period of at least 6 months. We also suggest using the body mass index (BMI) cut off values for age and sex to define overweight (25–30 kg/m²) or obese (≥ 30 kg/m²) as recommended by the International Obesity Taskforce [24]. For all anthropometric analysis, national reference values should be applied, or if not available the World Health Organization (WHO) standards should be applied (https://​www.​who.​int/​tools/​child-growth-standards/​standards).

Periorbital edema is the leading clinical sign of NS in children with a typical presentation. It may be asymmetrical initially and is frequently misdiagnosed as allergy. Edema is gravity-dependent, localized to the lower extremities in the upright position, and to the eyelids and the dorsal part of the body in a reclining position. The edema is painless, soft and pitting, keeping the marks of clothes or finger pressure. Anasarca may develop with ascites, and pleural and pericardial effusions. Efforts are underway to standardize the assessment of edema. Complications of NS may be the presenting symptoms or signs of the disease (e.g., abdominal pain related to severe hypovolemia, ascites, peritonitis, or pneumonia, dyspnea as a consequence of pleural effusion, ascites, pneumonia, or pulmonary embolism).

Extrarenal causes of edema should be considered including hepatic (hepatocellular insufficiency, cirrhosis, Budd-Chiari syndrome), digestive (exudative enteropathy, coeliac disease, lymphangiectasis), severe malnutrition, heart failure, hereditary angioneurotic edema, capillary leak syndrome, and thyroid abnormalities.

The diagnostic laboratory finding in children with NS is nephrotic range proteinuria (Table 1) defined by 3 + on urine dipstick in a spot urine, a urinary protein creatinine ratio (UPCR) ≥ 200 mg/mmol (≥ 2 mg/mg) or proteinuria > 40 mg/m²/h or ≥ 1000 mg/m²/day in a 24-h urine collection (Table 1). The use of a spot urine may be preferred to avoid sampling error and because of its excellent correlation with 24-h urine protein excretion [25]. Although urinary dipstick analysis is useful for screening and home monitoring, we recommend confirming nephrotic range proteinuria at least once by quantification of proteinuria either by spot urine sampling (if possible, first-morning void) or on a 24-h sample before initiating treatment for the first episode. First morning urine samples help rule out orthostatic proteinuria during follow-up to diagnose relapses [25, 26]. Typical semiquantitative dipstick results are shown in Supplementary Table S11. UPCr is preferentially used in SSNS as the urinary albumin creatinine ratio, although more specific, is less relevant in nephrotic range proteinuria. In addition, there are no universally accepted definitions for nephrotic range proteinuria when using urinary albumin creatinine ratio.

A physical examination for extrarenal features suggestive of genetic conditions is recommended (Table 2). Patients with extrarenal features suggestive of monogenic SRNS should primarily undergo genetic testing. Diagnostic work-up in patients with congenital NS (age < 3 months) should be done according to recent clinical practice recommendations [27, 28]. After the neonatal period, if family history is positive for SSNS, PDN therapy should be started as per this SSNS guideline. If family history is positive for a monogenic cause of SRNS, we recommend primary genetic testing.

In children, NS with onset at age above 1 year and typical presentation is most often SSNS associated with MCD. The likelihood of MCD is highest between ages 2 and 7 and decreases thereafter [9, 29]. Kidney biopsy allows the exclusion of the differential diagnoses (e.g., membranous nephropathy) and the confirmation of a primary podocytopathy (MCD, FSGS, or diffuse mesangial sclerosis (DMS)). Findings of DMS or membranous nephropathy have therapeutic implications as these entities are treated with specific protocols (membranous nephropathy) or may require genetic testing (DMS). Moreover, it allows the detection and grading of tubular atrophy, interstitial fibrosis, and glomerulosclerosis as prognostic markers [9].

However, there is not enough evidence to identify a clear age limit above which the probability is high enough for non-MCD pathology (e.g., membranous nephropathy), and thus the need for a kidney biopsy in children with NS. Therefore, we suggest that the decision of performing a kidney biopsy in older children (> 12 years) be made on a case-by-case basis. Atypical features suggesting the need for a kidney biopsy include macroscopic hematuria, low C3 levels, sustained hypertension, low estimated glomerular filtration rate (eGFR) not related to hypovolemia, arthritis and/or rash, or other extrarenal findings suggesting glomerulonephritis.

We also suggest a kidney biopsy be performed in patients with nephrotic syndrome and persistent microscopic hematuria in populations with a high incidence of glomerular diseases such as IgA nephropathy in East Asia. To reduce unnecessary kidney biopsies, the finding of more than 30 RBCs/HPF of fresh voided urine may be used as a criterion for performing a kidney biopsy in clinical practice [30].

About 50% of children with infantile onset NS (age 3–12 months) have a genetic cause of NS which usually does not respond to PDN treatment [31, 32]. The finding of DMS on kidney biopsy is highly suggestive for an underlying genetic defect, i.e., pathogenic variants in WT1, PLCE1, or PDSS2 genes [33‐36]. Therefore, we suggest following one of three strategies for infantile NS without extrarenal manifestations (Fig. 2): (i) primary genetic testing, if the results are rapidly available, with standard PDN treatment given if genetic testing is negative; (ii) primary kidney biopsy, followed by standard PDN treatment in the case of MCD and FSGS, genetic testing in the case of DMS, and specific treatment in the case of other underlying kidney histopathologies; and (iii) starting standard PDN treatment and then initiating genetic testing and kidney biopsy in case of SRNS.

SSNS follows a chronic course in most children and ideally all children with SSNS should be cared for by or in conjunction with a pediatric nephrologist at the outset. In some countries, the scarcity of pediatric nephrologists or the distance from tertiary referral centres, require general pediatricians to take primary responsibility [37].

Glucocorticoids are widely used in the treatment of NS, and their efficacy is well-established in children > 1 year of age with a typical presentation. In children between 3 and 12 months of age at onset, there is no evidence-based clear-cut approach to management. The management approach should consider the availability of time-sensitive genetic testing. In the absence of extrarenal features, priority may be given either to genetic testing, kidney biopsy or starting PDN, and assessing at 4 weeks (vide supra) (Fig. 2).

Because approximately 50% of children will develop FRNS or SDNS, the use of PDN in longer initial courses has been extensively studied for its efficacy to reduce relapses (Supplementary Table S3). Contrary to earlier evidence suggesting a benefit of longer courses of PDN [38], four recently published well-designed RCTs at low risk of bias, which evaluated 775 children, demonstrated that prolongation of PDN therapy beyond 2 or 3 months in the initial episode of SSNS does not reduce the risk of relapse [39‐42]. Since there are no adequately powered well-designed RCTs comparing 2 months with 3 months of PDN therapy, we recommend either an 8-week or a 12-week course of treatment of the initial episode of SSNS in line with KDIGO [15, 20] (Supplementary Table S3). The recent PREDNOS 2019 identified no differences in behavioral effects between different treatment durations [42]. Based on the available evidence, we recommend single daily PDN dosing.

Adverse effects of PDN in children with SSNS are common. An analysis of the adverse effects with PDN in 14 RCTs evaluating PDN therapy in the initial episode of SSNS with observation periods of 12–24 months found that hypertension (13%), psychological disorders (21%), cushingoid appearance (41%), and infections (22%) were common regardless of the total PDN induction dose used [10] (Supplementary Table S4). Future research recommendations are given in Supplementary Table S12.

Two small RCTs [43, 44] and one observational study [45] have demonstrated no differences in efficacy with a lower toxicity profile when PDN is administered as a single morning dose rather than divided doses. The potential benefits of the single-daily dose regimen include better adherence to therapy, lesser risk of hypothalamic–pituitary–adrenal (HPA) axis suppression and sleep disturbances. Dividing the dose has some practical considerations for medication use in children by minimizing the number of pills or volume of the liquid with each dose.

We do not recommend a tapering schedule during alternate day dosing. None of the four RCTs cited above used a tapering schedule of PDN in the experimental arm. Of the 775 children enrolled, there was only one possible case of adrenal suppression and that occurred in the control arm [41]. Treatment regimens in these four RCTs are shown in Table 3.

The traditional dose of PDN for induction of remission during the first episode of NS is 60 mg/m² per day or 2 mg/kg per day. Most country-based or international guidelines [15, 46‐48] recommend a maximum dose of 60 mg/day though the German guidelines recommend a maximum dose of 80 mg/day [46, 49]. No studies have formally evaluated the efficacy of doses higher than 60 or 80 mg/day in SSNS.

Although lower doses of PDN are associated with reduced risk of side effects, these doses may not be as effective. A single small RCT (n = 60) showed that a lower dose of PDN (40 mg/m²/day) during the initial episode of NS was associated with a longer time to remission compared to the standard dose (60 mg/m² per day; 11.4 ± 4.0 vs. 9.6 ± 2.6 days) [50]. At 24 months, the sustained remission rate was lower in boys receiving 40 mg/m² per day but there was no difference in girls [51]. A retrospective cohort of children with SSNS demonstrated that a lower cumulative dose of PDN (< 2500 mg/m²) used during the induction therapy for the first episode of NS is associated with shorter time to first relapse, higher rate of relapses and higher use of steroid-sparing agents, compared to higher doses (> 3000 mg/m²) [52]. Therefore, we recommend treating the first episode of NS with a dose of 60 mg/m² per day (or 2 mg/kg per day).

Younger children in particular will receive higher mg of PDN (up to 15% [53]) using a body surface area (BSA) compared to weight per kilogram dosing strategy. Limited knowledge exists regarding whether PDN dose should be calculated by weight or BSA. To avoid PDN overdosing in fluid-overloaded children, we suggest calculating the PDN dose based on the estimated dry weight. Two small RCTs [54, 55] with 146 participants compared weight-based dosing with BSA-based dosing in young children (weight < 30 kg, BSA < 1 m²) with their initial episode of SSNS and with relapse of SSNS. There were no statistically significant differences for efficacy or steroid toxicity when comparing weight-based versus BSA-based dosing of PDN but follow-up duration was short in both studies. One patient in the BSA group developed hypertensive encephalopathy [55]. Mean cumulative PDN dose was lower with weight-based dosing in both studies [54, 55]. When height is not available, PDN doses which approximate to 60 mg/m² and 40 mg/m² can be estimated from the formulae: 2 × weight + 8 and weight + 11, respectively [56].

Studies aiming to reduce the number of relapses by adding a non-glucocorticoid immunosuppressive (steroid-sparing) agent to PDN therapy for the initial episode of NS are scarce. Zhang et al. studied the efficacy of adding azithromycin in combination with PDN therapy in children with their first presentation of NS [57]. The median duration before remission was 6 days in the group that received azithromycin in addition to PDN, and 9 days in the PDN alone group (p < 0.0001). There were no differences in terms of relapses at 6 months.

An RCT demonstrated that adding 8 weeks of cyclosporine (CsA) to PDN within the first 4 weeks of treatment of the first episode of NS (after establishing remission over 3 days) reduced the risk of first relapse within the first 6 months (RR 0.33, 95% CI 0.13–0.83), but no difference was observed at 12 months (RR 0.72, 95% CI 0.46–1.13) [58]. There are RCTs in progress in children studying the benefits of adding mycophenolate mofetil (MMF) [59] or levamisole (LEV) [60] to PDN during the initial episode of NS, as soon as children have entered remission, but there are no published results to inform the guideline. Moreover, a significant percentage of children with SSNS are infrequent relapsers and will never require a steroid-sparing agent. Therefore, due to the potential unnecessary side effects and to added cost, initial therapy combining steroids and a steroid-sparing agent cannot be currently recommended.

For the management of childhood NS, both prednisone and prednisolone have been used interchangeably, and at an equivalent dose. Prednisone is a prodrug of prednisolone [61]. The conversion of prednisone to the biologically active prednisolone occurs mainly in the liver. This interconversion is not a limiting factor, even in patients with severely impaired liver function [62, 63]. NS does not influence the conversion of prednisone to prednisolone [64, 65]. Acute NS and the hypoalbuminemic state do not reduce absorption of PDN or the conversion of prednisone to prednisolone [65, 66]. In clinical practice, prednisolone and prednisone are usually given orally. Prednisolone is palatable and is the preferred choice for young children [67, 68].

Deflazacort is a synthetic glucocorticoid oxazoline derivative of prednisolone. Six milligrams of deflazacort have approximately the same anti-inflammatory potency as 5 mg of prednisolone or prednisone. There was no difference between deflazacort and PDN in the number achieving remission in the first episode of SSNS in two small RCTs [69, 70]. However, fewer children relapsed following deflazacort treatment compared with PDN [69, 71]. There is a report of toxic epidermal necrolysis in two children with NS who received deflazacort [72]. At this time, there are insufficient data to recommend the use of deflazacort rather than PDN in the treatment of NS.

Intravenous methylprednisolone at equivalent doses of oral prednisone (equivalent dose is 5 mg for every 4 mg of IV methylprednisolone) may be used in situations where a patient is unable to tolerate oral medications or if adherence may be a problem. Intravenous therapy should be limited to a short duration with the intent to switch back to oral medication at the earliest opportunity.

The PREDNOS 2 RCT [82], which was adequately powered, generalizable to the overall SSNS population, and at low risk of bias, evaluated 271 children with NS and URTI. The study found no benefit of administering five days of low dose PDN (15 mg per m² BSA which is equivalent to 0.5 mg/kg) at the onset of URTI in preventing relapse. The finding was consistent among subgroups of children receiving alternate day PDN or children receiving alternate day PDN and other immunosuppressive agents, although the study was powered for whole group analysis only. In contrast, four smaller RCTs [83‐86] including between 36 and 89 patients, reported that using low dose daily PDN at the onset of a URTI reduced the number of children with a subsequent relapse. These four studies were all at high risk of bias for one or more study attributes and were conducted in different geographic regions as compared to the low risk of bias study. Poorly designed RCTs at increased risk of bias are more likely to overestimate the efficacy of a treatment due to confounding, and/or selective or underreporting of outcomes in treatment groups [87, 88]. The baseline risk of an URTI triggering a relapse determines the number needed to treat to prevent one relapse with the intervention. Within most of the studies considered here [83‐86] and in a demographic study [89], the risk is approximately 50%, but it was much lower (20%) in PREDNOS 2. Overall, there is insufficient evidence to recommend the routine use of a short course of low-dose daily PDN at the onset of an URTI for prevention of relapses. However, such an approach may be considered in children already taking low-dose alternate day PDN and at a greater risk of URTI triggering relapse. A cost-effectiveness analysis of PREDNOS 2 showed giving daily oral PDN to be dominant in health economic terms [90]. This was due to a small cost benefit driven largely by the low-cost of PDN, and reduced health-related quality-of-life associated with a relapse for the small (but clinically non-significant) additional number of children who relapsed in the placebo arm [90]. (Further information is given in Supplementary Table S5).

SSNS is a relapsing–remitting condition. Children with frequent relapses, who require frequent courses of oral PDN, particularly in the presence of comorbidities, may develop steroid toxicity (Table 5). In children with FRNS or SDNS, it is necessary to balance risks and benefits of the intervention on an individual basis. The objective is to keep each patient controlled on therapy with minimal adverse effects. In some centers, the initial approach in children with FRNS is low-dose maintenance oral PDN, while in other centers a steroid-sparing agent is immediately started.

The use of low-dose PDN in children with FRNS to maintain remission is primarily based on two historic small single-arm, uncontrolled studies with alternate day [91] or daily dosing [92]. Alternate-day dosing has been more widely adopted, although this is not evidence-based. A single open-label RCT [93] involving 61 patients with FRNS found that low dose daily (0.25 mg/kg) compared with alternate-day (0.5 mg/kg) PDN reduced the risk for relapse during 12 months of follow-up (0.55 relapses/person-year compared with 1.94 relapses/person-year) and lowered one year of PDN exposure (0.27 ± 0.07 versus 0.39 ± 0.19 mg/kg/day) with no differences in adverse effects. There was some clinical evidence of reduced glucocorticoid toxicity with the daily dosing schedule. The preferred use of daily or alternate-day low dose PDN for relapse prevention in FRNS requires additional study. Transition to steroid-sparing agents is recommended in patients not controlled on therapy as defined in Table 1.

Steroid-sparing agents used in children with SSNS include CNIs (cyclosporin A (CsA), tacrolimus (TAC)), cyclophosphamide (CYC), immune modulators (levamisole (LEV), anti-proliferative agents (mycophenolate mofetil (MMF)/mycophenolic sodium (MPS)), and anti-CD20 monoclonal antibodies, primarily rituximab (RTX). There is insufficient evidence to establish the best initial option and the optimal sequence of agents from least to most effective or least to most toxic. The choice of agent should be based on family and physician preferences and the risk profile for drug-associated complications. Factors to consider include disease type/severity, age—including onset of puberty, potential adherence, side-effect profile, comorbidities, cost and availability. In the following sections, we discuss the pros and cons of each available agent and provide a roadmap, based on the available evidence, of reasonable choices based on the clinical features of each patient with SSNS. Regarding a switch from one steroid-sparing agent to another, the same considerations apply. Moreover, we have added the definition of “controlled on therapy” to provide a timeframe for this decision.

In Table 5, we provide dose, monitoring, adverse effects, and considerations on cost for therapeutic agents that are currently used for relapsing SSNS patients. In Supplementary Table S6, we provide GRADE-based evidence, given the available RCTs (Supplementary Table S7), on the different steroid-sparing therapeutic agents. An overview of recent observational studies on steroid-sparing therapeutic agents is given in Supplementary Table S8.

CNIs have been used to treat relapsing SSNS for nearly 30 years [94‐98]. Because of the lack of cosmetic side effects, TAC may be preferred to CsA. A Cochrane systematic review did not identify any RCTs comparing CsA with TAC in children with SSNS [12]. In Japan, an RCT comparing TAC and CsA is currently underway (jRCTs031180132, UMIN000004204).

CNIs are effective in maintaining remission in children with FRNS and SDNS. A single RCT performed in Japan and including 108 children with FRNS/SDNS demonstrated that CsA compared with placebo reduced the risk of relapse (relapse rate ratio 0.55 (95% CI 0.37–0.82)) [99]. Observational studies have also demonstrated reduced relapse rates with CsA compared with PDN [95, 100‐105]. However, many patients suffer relapses when CsA is ceased [101‐104, 106]. Ishikura et al. reported that 84.7% of patients had a relapse within 2 years after completion of the 2-year CsA therapy and 59.2% of patients had regression to FRNS [106]. There are small RCTs comparing alkylating agents or MMF with CsA. Compared with alkylating agents, the number of patients relapsing by the end of therapy (6–9 months) on CsA may not differ (2 studies, 95 children: RR 0.91, 95% CI 0.55 to 1.48). However, following cessation of these medications and because the effect of alkylating agents but not CsA is prolonged after cessation, fewer children relapse after receiving alkylating agents compared with CsA alone (risk of relapse at 12–24 months; 2 studies, 95 children: RR 0.51, 95% CI 0.35 to 0.74) [12].

Two small RCTs suggested that the number of patients relapsing by 12 months may not differ between MMF and CsA (2 studies, 82 children: RR 1.90, 95% CI 0.66 to 5.46) but there is considerable imprecision in these findings. The addition of a third study to the meta-analysis indicated that the relapse rate/year may be higher with MMF than with CsA (mean difference 0.83 (95% CI 0.33 to 1.33) [12].

In RCTs, MMF is less likely to cause hypertrichosis and gum hypertrophy compared with CsA [12, 107‐109] but no differences in other adverse effects (hypertension, impaired kidney function and infections) were identified. Three large observational studies [14, 110, 111] found higher efficacy in maintaining remission with CNIs compared with MMF. However, adverse effects were more common with CNIs.

The use of TAC in SSNS is based on the effectiveness of CsA in SSNS [95], on the results of observational studies [14, 97, 110] and the efficacy of TAC in pediatric kidney transplantation.

There are no RCTs that compare TAC to CsA. A trial of TAC versus CsA for FRNS in children is being conducted in Japan (jRCTs031180132, UMIN000004204). Only small-number case series are available [98, 112‐114]. Switching from CsA to TAC is only effective in reducing cosmetic side effects but warrants caution for the potential onset of diabetes mellitus [114].

Patients should be monitored for side effects as indicated in Tables 4 and 5. Therapeutic drug monitoring indications are given below.

Nephrotoxicity is the most problematic side effect of CsA, and its risk is increased after use for > 2 years [115, 116]. CsA-induced chronic nephrotoxicity cannot be diagnosed based only on urinalysis or blood tests. It is advisable to avoid prolonged use of CsA and to consider its discontinuation or to perform a kidney biopsy after 2–3 years to avoid/detect toxicity. However, there is no definitive evidence supporting the necessity of kidney biopsy in SSNS treated with CNIs. Recent clinical studies of micro emulsified CsA [100, 117] have demonstrated a lower incidence of nephrotoxicity.

Cosmetic side effects, such as hypertrichosis and gum hyperplasia, are common with CsA [100‐105]. Infections, hypertension, and posterior reversible encephalopathy syndrome (PRES) are also known complications of CsA therapy [100‐105, 118].

Among the side effects of TAC, new-onset diabetes mellitus is important. Particular caution is required when TAC is used in patients with a family history of diabetes mellitus or if risk factors for impaired glucose tolerance (e.g., obesity) are present [119]. Renal interstitial fibrosis has also been reported, as with CsA. One report described a significant association between higher TAC trough levels and renal interstitial fibrosis [112].

The dose of CsA should be adjusted with drug monitoring based on assays validated against tandem mass spectrometry. According to a multicenter, prospective RCT of Sandimmun® conducted in Japan on 44 children with FRNS, sustained remission was significantly higher in the dose-adjusted group (initially the dose was adjusted to maintain blood trough levels within 80–100 ng/mL for the first 6 months, and then within 60–80 ng/mL for the next 18 months) compared with the 2.5 mg/kg fixed-dose group (initially the dose was adjusted to maintain blood trough levels within 80–100 ng/mL for the first 6 months, but then fixed at 2.5 mg/kg for the next 18 months) (50 vs. 15%; p < 0.01) [95]. A multicenter observational study assessed Neoral® [101], a microemulsified preparation of CsA, in 62 children with FRNS, with adjustment of the dose using the same target trough levels as stated above. This study reported that microemulsified CsA was effective and safe (relapse-free survival rate at month 24, 58%; incidence of nephrotoxicity, 8.6%), similar to conventional CsA [100].

The AUC_0–4 (area under the time-concentration curve) of CsA is best predicted by C₂ (CsA blood concentration at 2 h post-dose) in kidney transplant patients [120]. Similar findings were reported in children with NS [121]. A multicenter, prospective, RCT in Japan on 93 children with FRNS compared two different target C₂ levels: a higher C₂ group (target C₂ 600–700 ng/mL for the first 6 months, followed by 450–550 ng/mL for the next 18 months) and a lower C₂ group (target C₂ 450–550 ng/mL for the first 6 months, followed by 300–400 ng/mL for the next 18 months) [94]. At 24 months, the relapse rate was significantly lower in the higher C2 group than the lower C₂ group (0.41 vs. 0.95 times/person-year; hazard ratio, 0.43; 95% confidence interval, 0.19 to 0.84; p < 0.05). The rate and severity of adverse events were similar in both treatment groups [94].

Absorption of oral CsA after pre-meal administration (15–30 min prior to a meal) is greater than post-meal administration so it may be preferable to administer CsA before meals. The main priority is to give it in a consistent manner. Concomitant use with other drugs requires adequate attention since macrolide antimicrobials and many other drugs can affect metabolism. Grapefruit juice should be avoided as it inhibits metabolism of CsA and causes increased blood concentrations of the drug.

Tacrolimus requires adjustment of dosage by monitoring blood concentration. However, safe and effective dosage and mode of administration of TAC have not yet been established in children with SSNS. Suggested dosage and blood levels are extrapolated from data on kidney transplant recipients.

CsA is very effective in the treatment of FRNS/SDNS and allows steroid tapering and discontinuation in the majority of patients [95, 100‐105]. The shortcoming of CsA therapy is that many patients experience relapse after termination of CsA therapy (CsA dependence) [101‐104, 106]. Moreover, CNIs have a variety of side effects, including nephrotoxicity. In comparison to CsA, TAC has fewer cosmetic side effects.

If a child remains in sustained remission for at least 12–24 months and off steroids, CNI discontinuation should be considered to avoid nephrotoxicity [115, 116]. Tapering CNI dose to zero over about 3 months rather than discontinuing abruptly may be preferable because in case of a reappearance of proteinuria during tapering, reestablishing the initial CNI dose may be sufficient to avoid a relapse and a course of oral PDN while establishing that the patient still needs maintenance therapy.

A meta-analysis of 4 RCTs with 161 participants [12] comparing CYC with PDN or placebo showed a reduction in the number of relapses by 6 to 12 months (4 studies, 161 children; RR 0.47 [95% CI 0.34, 0.66]) [12]. A single course of monthly intravenous doses of CYC at a dose of 500 mg/m² per dose (max single dose 1 g) × 6 months can be given when adherence is an issue [122, 123].

A review of 38 RCTs and observational studies assessing alkylating agents (CYC and chlorambucil) [13] including 1504 patients and 1573 courses and published between 1960 and 2000, indicated sustained remission rates of 72% after 2 years and 36% after 5 years for FRNS; the rates were 40% and 24%, for SDNS respectively. The maintenance of sustained remission declines with time, i.e., 44–57% at 1 year, 28–42% at 2 years, 13–31% at 5 years [124‐128]. The effect may be lower in children below 3–5.5 years of age [125, 127, 129].

In comparison with CsA courses limited to 6–12 months (two RCTs), the actual percentage of sustained remission at 2 years for alkylating agents was higher, indicating that the effect of alkylating agents lasted longer than CsA after cessation of therapy [12]. One non-randomized comparator trial ([130], n = 46) suggests that RTX is non-inferior to CYC in maintaining remission over 1 year.

Cyclophosphamide treatment should be initiated after the patient has achieved remission and has been treated with the recommended dose of PDN for relapse. Published literature examining the use of CYC does not directly address whether co-intervention with PDN is necessary to reduce relapses or risk of adverse effects. Descriptions of continuation of PDN or concomitant administration of PDN while on CYC vary widely in the literature. Protocols ranged from PDN 10–40 mg/m² either daily or alternate days, to 60 mg/m² every other day. Tapering at the end of treatment was also highly variable [13, 96, 124, 131]. Due to substantial variation in practice, administering CYC in combination with alternate-day oral PDN starting with a dose of 40 mg/m² (1.5 mg/kg) and reducing to 10 mg/m² (0.3 mg/kg) over the duration of treatment was considered as reasonable practice by the guideline committee. Alternate-day oral PDN may help to reduce the risk of neutropenia when starting CYC initially.

Leukopenia occurred in 32.4% of patients on CYC and was more common with CYC alone than with CYC plus PDN protocols (22/38 vs. 8/52) [13]. The Latta meta-analysis reported reversible alopecia in 17.8%, infections in 1.5%, hemorrhagic cystitis in 2.2%, and malignancy in 0.2%. However, the cumulative dose used in many of the included studies was higher than current recommendations [13]. Studies using lower cumulative doses [124, 132] report transient leukopenia (7 to 23%) as the main adverse effect with transient alopecia and hemorrhagic cystitis occurring in ≤ 1%. However, long-term follow-up studies in patients who have been treated with these lower doses are lacking.

Alkylating agents, in particular CYC, have been used in pediatric NS for over 5 decades. Considering that other alkylating agents, i.e., chlorambucil, are currently rarely used for children with SSNS and showed a worse safety profile compared to CYC [13], we have focused on CYC. CYC is relatively inexpensive and monitoring requirements involve relatively inexpensive and readily available standard tests. Compared to agents like LEV, MMF and CNI, CYC is administered for a short-term course with a sustained effect after discontinuation. Thus, safety monitoring is needed for a shorter duration. The risk of gonadal toxicity is reduced with appropriate restriction of total cumulative dose. CYC should be used with caution in peri-pubertal males. The risk of hemorrhagic cystitis is very low with oral therapy at this recommended dose and with maintenance of fluid intake and diuresis. Leukopenia/neutropenia is the most common AE expected, and dose adjustment is a component of all protocols. Of note, CYC’s use requires additional treatment with oral PDN, which may promote further steroid toxicity. On balance, the potential risks of CYC may favor the use of other steroid-sparing agents, if available.

A recent international multicentre RCT has enhanced the quality of evidence for the effectiveness and safety of LEV. Gruppen (2018) [139] compared LEV therapy to placebo in 99 children with FRNS or SDNS and found a significant reduction in the number of relapses at 12 months (RR of relapses on LEV 0.77, 95% CI 0.61 to 0.97) [12]. Thus, 26% of children in the LEV group compared with 6% in the placebo group remained in remission at 12 months. Eight RCTs (474 participants) combined in a meta-analysis [12] indicated a benefit of LEV over PDN, placebo or no treatment (RR 0.52, 95% CI 0.33 to 0.82).

Small comparative RCTs comparing LEV with CYC [140, 141] showed no difference in efficacy but were not powered to show a difference. An RCT found no difference in efficacy between MMF and LEV but MMF levels were not measured [142]. The Gruppen 2018 [139] and Sinha 2019 [142] studies suggest that LEV may be more effective in FRNS than SDNS. These recent RCTs [139, 142] used a dose of LEV of 2.5 mg/kg/alternate day, maximum 150 mg, for 12 months. Most other recent studies used doses of 2–3 mg/kg on alternate days for 6–24 months. Some observational studies have used doses of 2–2.5 mg/kg daily for 4–24 months [143‐149] with three studies [147‐149] suggesting reductions in relapse rates in patients who had not responded to alternate-day LEV. These data require further larger RCTs, powered to detect a difference, if any, for confirmation.

Common adverse effects include rashes, leukopenia, and abnormal liver function tests. These are generally transient and reversible on discontinuation of therapy. Rarely ANCA positive arthritis (2% in Gruppen 2018 [139]), rash and other vasculitis symptoms have been reported, which resolve upon LEV discontinuation.

While most adverse effects are transient and reversible on discontinuation, the main emerging threat is ANCA-positive vasculitis particularly with prolonged use. Regular monitoring as indicated in Tables 4 and 5 is advised with cessation of therapy if ANCA titers are positive.

Available studies do not comment on this. Discontinuation without tapering should be considered once the patient is in sustained remission and off steroids for at least 12 months.

LEV is an immunomodulant which has been used for over 3 decades in NS. Its low cost makes it a useful option, particularly in low resource settings. However, it is unavailable in some countries. Lack of nephrotoxicity and ease of monitoring are other major advantages. When introducing this agent, some physicians prefer to maintain low-dose alternate-day PDN on non-LEV days for a few months, then oral PDN is tapered and stopped, and the patient remains on LEV alone.

The standard dose for MMF in RCTs is 1200 mg/m²/day divided into two doses every 12 h orally with a maximum daily dose of 3000 mg. Five hundred mg of MMF corresponds to 360 mg of MPS. Patients may be started on half dose and dosage may be increased after 1 week in case of no side effects, e.g., leukopenia or GI discomfort.

Patients should be monitored for side effects as indicated in Tables 4 and 5. Therapeutic drug monitoring indications are given below.

Assessment of mycophenolic acid (MPA) trough levels is not recommended as there is a poor correlation with efficacy and safety using single pre-dose measurements [150, 151]. A limited sampling strategy for assessing pharmacokinetic profiles was established in children with NS on MMF monotherapy being in remission [152], whereas such a profile is not available for those on MPS. It requires three measurements of plasma MPA at times 0 min (before administration, C₀), 60 min (C₁), and 120 min (C₂) after administration, and allows a good estimation of MPA-AUC_0-12 using the formula eMPA − AUC₀₋₁₂ = 8.70 + 4.63 * C₀ + 1.90 * C₁ + 1.52 * C₂ [152]. In children with FRNS with MPA AUC_0-12 > 50 mg × h/L estimated using the formula eMPA—AUC = 7.75 + (6.49 * C₀) + (0.76 * C_0.5) + (2.43 * C₂) [108, 153], the efficacy of MMF was similar to that of CsA [108]. The latter formula was originally established in adult heart transplant patients treated with concomitant CsA. We recommend using therapeutic drug monitoring in patients not controlled on MMF therapy despite adequate dosing aiming for eMPA-AUC_0-12 > 50 mg × h/L. For this purpose, either one of the above mentioned formulas can be used [108, 152, 153]. It should be noted that immunoassays for the determination of MPA plasma levels measure 10–20% higher MPA plasma levels than high-performance liquid chromatography (HPLC) or mass spectrometry (MS) due to cross-reactivity with MPA metabolites [154, 155].

No RCTs have compared MMF or MPS with PDN in children with FRNS or SDNS. However, numerous observational studies [156‐160] (Supplementary Table S8) have reported that MMF or MPS are more effective than PDN in maintaining remission in children with FRNS or SDNS. These studies showed an approximately 50% reduction in the relapse rate on MMF/MPS, enabling reduction in dose or cessation of PDN. Studies have not specifically compared the relative efficacies of MMF/MPS in children with FRNS or SDNS.

Four RCTs compared MMF with other steroid-sparing agents in FRNS and SDNS. Three RCTs compared MMF with CsA in 142 children. Two RCTs [107, 108] combined in meta-analysis found no difference in the number of children with relapse between MMF and CsA (82 children: RR 1.90, 95% CI 0.66 to 5.46) [12]. However, the relapse rate/year was higher in children treated with MMF compared with CsA (3 studies, 142 children: mean difference 0.83, 95% CI 0.33 to 1.33) when a third study was included [12]. One RCT compared MMF with LEV and found no difference in the number of children with relapse at 12 months between treatments [142]. MPA levels were not measured in this study.

Three observational studies involving 312 children with FRNS or SDNS compared MMF with TAC [14, 110] or CsA [111]. MPA levels were not monitored in these studies. Two of these studies [14, 111] found better efficacy for maintaining remission with CNIs compared with MMF though adverse effects were more common with CNIs.

The most common adverse effects of MMF are abdominal pain, loss of appetite, diarrhea, and weight loss. This is less likely to occur with enteric-coated MPS. However, some individuals tolerate MMF better than MPS. Other adverse effects are leukopenia, anemia and elevated hepatic transaminases. These adverse effects are uncommon and usually mild. Monitoring for side effects should be done as indicated in Tables 4 and 5. MMF/MPS is teratogenic in the early months of pregnancy so effective contraception should be used by all sexually active female adolescents during MMF/MPS therapy. In males, recent evidence in patients receiving MMF/MPS after kidney transplantation and a large meta-analysis of different drugs [161] indicates that the risk of congenital malformations is comparable to that of the general population [162].

There is now extensive documentation of the successful and safe use of MMF in children with FRNS and SDNS but studies did not differentiate between these groups. In clinical practice, MMF appears more effective in children with FRNS. Its advantages consist in lack of nephrotoxicity and of cosmetic side effects compared to CNIs.

There are no studies on the duration of MMF/MPS use or on when to discontinue MMF/MPS. If the child achieves control on therapy for at least 12 months, then consideration may be given to tapering MMF over 3–6 months and then discontinuing it. As with CNIs, the advantage of tapering over abrupt discontinuation is that in case of proteinuria, re-establishment of MMF at initial dose may be sufficient to avoid a relapse while establishing that the child still requires maintenance treatment. The use of more extended periods may be considered, especially in peri-pubertal age or in the presence of previous severe steroid toxicity.

In terms of dosing regimen, the original course of RTX used for lymphoma patients required 375 mg/m² given as an IV infusion weekly for 4 doses. The RTX protocols used in the available RCTs and observational studies in children with FRNS/SDNS included single, 2, 4, and 7 infusions. In addition to variability in the number of RTX infusions, there has been variation in RTX dosing, ranging from 375 to 1500 mg/m² per treatment, although most studies used 375 mg/m². The dose of 750 mg/m² has not been associated with a better response rate than 375 mg/m²; however, a lower dose of RTX (100 mg/m²) has been associated with the risk of earlier relapse (reviewed in [163] and in [164]). In terms of infusion number per course of RTX treatment, the use of a single infusion at standard dose followed by monitoring of CD19 ( +) cells at 7 days is derived from studies performed in adults with ANCA-associated renal vasculitis and membranous nephropathy. If at 7 days post-infusion, the percentage of total B cells is < 1% of total lymphocytes this indicates adequate B cell depletion [165]. Reconstitution of B cells is defined when total B cell counts are > 5/mm³ in absolute number [166].

During the last decade, a number of RCTs have shown that RTX is reasonably safe in the short term and relatively effective when compared to other immunosuppressants as a steroid-sparing treatment. However, studies differ in terms of populations, number of doses of RTX, additional medications and comparators. Unlike other immunosuppressants, the lack of long-term follow-up in RTX-treated patients must be considered at the time of clinical decision.

Eight RCTs have evaluated the efficacy of RTX in children with FRNS or SDNS. Four RCTs evaluated 1 to 4 doses of RTX in children with SDNS and CNI dependence compared with placebo [167, 168] or CNIs [169, 170]. Four studies compared 1 to 2 doses of RTX in children with SDNS or FRNS on low-dose PDN compared with TAC [171], low dose PDN [172, 173], or low dose MMF [174]. A meta-analysis showed that the number of patients with relapse fell by 80% by 6 months and 50% by 12 months after treatment [12]. Longer duration of remission was seen in children whose relapses were previously managed with PDN alone [172, 173]. Moreover, a large retrospective study assessing RTX use in more than 500 children with FRNS/SDNS showed that patients were 19% more likely to relapse for each additional steroid-sparing agent received prior to RTX, and that younger age at first infusion was associated with earlier relapse [164, 175, 176].

Adverse events were generally limited to mild infusion reactions. There was no increase in infections. RTX-related neutropenia (RRN) has been well documented in the literature, although the exact mechanism is not well known. In children, RRN is usually not associated with serious bacterial or viral infections and most of the reported infections are self-limiting. Supplementation with granulocyte colony stimulating factor (G-CSF) may not be needed, especially in late onset neutropenia, i.e., neutropenia occurring 4 weeks after last RTX infusion [177‐179].

No deaths or serious adverse reactions were recorded in RCTs on the use of RTX in children with SSNS. While there are case reports of fatal lung fibrosis, immune-mediated ulcerative colitis, fulminant myocarditis, Pneumocystis jiroveci pneumonia following RTX use in children with SSNS, a retrospective survey of 511 children with SSNS and treated with RTX [180] identified only two children with life-threatening but non-fatal complications (Pneumocystis jiroveci pneumonia, myocarditis). However, prolonged and significant reduction of total memory and switched memory B cells together with hypogammaglobulinemia has been demonstrated in patients following RTX, particularly in young patients with SSNS [181].

Exclusion of certain infections and monitoring for side effects should be done as indicated in Tables 4 and 5.

RTX treatment has proven reasonably safe and effective for both FRNS and SDNS. Given its uncertain long-term safety profile, it is advisable to use RTX as a second-line steroid-sparing agent in children who are not controlled on therapy with a first-line steroid-sparing agent. Since long-term side effects such as hypogammaglobulinemia appear to be more likely and efficacy appears to be less convincing in younger children, the use of RTX may be reserved for older children.

Following the first course of RTX, diverse approaches to repeated courses have been proposed, based either on disease relapse, on B cell reconstitution or on time elapsed from the initial treatment. Evidence for the most correct approach is lacking [164]. Based on a recent retrospective survey, 30 of 346 included children tolerated up to 7 courses of RTX infusions (mainly dosed with 375 mg/m²/course) with an acceptable side effect profile (most common hypogammaglobulinema, followed by infections and neutropenia) and good efficacy [182].

It is unknown to what degree other immunosuppressive agents should be tapered or discontinued following RTX administration. In most studies, PDN at alternate-day doses was tapered off within 2 months before CNIs were reduced and stopped. If patients were taking MMF and mizoribine, these drugs were discontinued after the first dose of RTX. A recent study [180] demonstrated that treatment response depends on both RTX dose and on the use of maintenance immunosuppression. The study documented that in complicated FRNS and SDNS patients, giving “low dose”, i.e., 375 mg/m² RTX and maintaining immunosuppression (IS), most frequently with MMF but in some cases with either CNI or oral PDN, was equivalent in terms of median relapse-free period to giving higher doses without maintaining IS after RTX [180]. In SDNS, a small prospective cohort study found that relapse-free survival 12 months after RTX therapy was higher in children receiving MMF than in children not receiving MMF [183]. An RCT evaluating MMF post-RTX treatment in “complicated” FRNS and SDNS showed that this approach was helpful in preventing relapse in 80% of patients [166]. An RCT comparing maintenance MMF to repeated RTX infusions in children with SDNS is ongoing (RITURNS II Study, NCT03899103). The use of CNIs following RTX infusions may be equally helpful, but this has not been formally assessed. These data suggest that in children with SDNS not controlled on RTX alone, following subsequent RTX infusions, the strategy of maintaining an oral steroid-sparing agent (MMF or a CNI) for at least 6 months may promote sustained remission.

As with all steroid-sparing agents and even more with RTX given its long-lasting effect, once the child is controlled on therapy, RTX infusions should be discontinued.

In addition to RTX, other monoclonal antibodies targeting B cells or modulating their function or depleting plasma cells have been employed in the treatment of SSNS.

Ofatumumab, in contrast to rituximab, is a fully humanized anti-CD20 monoclonal antibody. A case report described two boys, aged 3 and 14 years, with persistent SSNS, who were allergic to rituximab. Both children achieved a prolonged remission exceeding 12 months following the administration of a single dose of ofatumumab [184]. However, a recent clinical trial comparing RTX and ofatumumab in randomized 140 children with SDNS and found that there was no difference in the percentage of participants who relapsed at 12 or 24 months [185].

The combination of different steroid-sparing agents is not supported by adequate evidence. There are no RCTs that compare the combination of CNI plus MMF vs. CNI or MMF alone. There is a single observational study involving 130 Pakistani children with SSNS. Of these 20 had suboptimal response to MMF and CsA was added. Nineteen out of 20 benefited but only 4 had CR and 9 were CNI-dependent. In a retrospective publication on the use of RTX [180], the prolonged use of MMF or other steroid-sparing agents following a single cycle of RTX was found to induce stable remission in those receiving low-dose RTX (375 mg/m² per course) but there was no increase in benefit in those receiving higher doses of RTX (750 mg/m² or higher). We suggest that if children with FRNS or SDNS are controlled on therapy with more than one immunosuppressant (i.e., steroid-sparing agent plus maintenance PDN or CNI plus MMF), discontinuation of the most toxic agent be implemented.

A single RCT found no definitive benefit of azithromycin compared with PDN in the initial episode of SSNS [57]. Single RCTs found no benefit of azathioprine, ACTH or mizoribine in children with FRNS/SDNS [186‐188].

Children in the acute nephrotic state are at increased risk for venous and arterial thromboembolic events that disappears when the child achieves remission. The clinical spectrum includes cerebral venous thrombosis, deep venous thrombosis, pulmonary embolism, and arterial infarction but the majority of children have deep venous thromboses rather than arterial thromboses [203, 204]. The reported incidence of symptomatic thromboembolic events, mainly diagnosed within 3 months after disease onset [204], is about 3% in all forms of NS with peaks in infancy and adolescence (summarized in [205]) and is much lower than in adults (27%). The incidence is lower in children with SSNS (1.5%) than in complicated NS/SRNS (3.8%) [206]. Associated risk factors include the disease-related hypercoagulability, hypovolemia, immobilization, infections with hospitalization, indwelling central venous lines, and underlying hereditary thrombotic predisposition [204, 207, 208].

There is insufficient evidence to recommend routine prophylactic anticoagulation during the acute nephrotic state in children and adolescents. It is essential to assess the individual clinical risk profile of each child by taking a detailed history of previous thromboembolic events and hereditary predisposition, evaluating the volume status, and avoiding iatrogenic thrombotic risk factors. If preventive anticoagulation is needed, based on the individual clinical risk profile, we suggest using low-molecular weight heparin [209]. There are insufficient data to give recommendations on the use of antiplatelet treatment with aspirin in children with NS.

Conflicting data have been published on the risk of glucocorticoid-induced osteoporosis (GIO) in pediatric SSNS. Some studies reported low bone mineral density (BMD), correlating with disease severity and cumulative steroid intake [254‐257]. In contrast, others have reported no change in BMD after initial, intermittent or long-term alternate-day therapy [258‐262]. Children and adolescents with FRNS/SDNS seem to be at a higher risk of developing low BMD [263, 264]. In summary, bone mineral loss may occur early with high-dose daily PDN (which is usually given at start of therapy) but is less significant with subsequent intermittent or low-dose alternate-day regimens. The reported incidence of fracture is low (6–8%) [263, 264]. No data are available on the use of biphosphonates in children with NS. The prevention or limitation of GIO by minimizing steroid exposure to the lowest dose and shortest effective regimen is recommended. Nutritional and lifestyle measures to maintain bone strength should also be continued.

Both the vitamin D-binding protein (VDBP) and albumin bound fractions of vitamin D are lost in urine in NS relapse, and several reports have documented low levels of total serum 25(OH)D in and after NS relapse [265‐267]. The total serum 25(OH)D levels were shown to return to levels similar to healthy controls after 3 months of attaining remission by Banerjee et al. [268], whereas two other studies reported persistent low 25(OH)D levels at 3 months [267, 269]. In contrast, the biologically active fraction of free 25(OH)D levels were found to be similar to levels in healthy children both in remission and relapse of NS [270].

In patients with SSNS on steroid therapy, there are conflicting results about improvement of BMD when treated with vitamin D and calcium [271‐274]. Calcium and vitamin D supplementation does not specifically treat GIO and there is insufficient evidence to recommend routine supplementation of vitamin D₃ and oral calcium at onset of SSNS or during relapses of usually short duration. However, ensuring adequate calcium intake and normal 25(OH)D serum levels is suggested to optimise bone health. Vitamin D supplementation should be guided by serum levels, checked after remission of at least 3 months, and by national pediatric guidelines for vitamin D deficiency [275]. Excess supplementation has been associated with hypercalciuria [274, 276]. Note that higher 25(OH)D target levels are recommended in children with CKD stages 2–5D [277].

Regular physical activity can prevent thrombosis and skeletal changes. Healthy nutrition is recommended and should be guided by a specialized dietician. Eating home-prepared meals using fresh ingredients instead of canned, frozen, or packaged meals is preferable, since the latter have a much higher salt content. As increased oral protein intake has not shown to improve serum albumin levels or patient outcomes, a regular oral protein intake is recommended [280].

Sun protection as a general supportive measure is important in all children, especially in those on long-term immunosuppression. Measures include reducing exposure to UV radiation, avoiding sunbathing, covering the skin with adequate clothing, and using sun protection creams with high to very high sun protection factor.

While children are less likely to relapse as they grow older [281], more than 10% (6.8–42.2%) of childhood-onset SSNS patients still experience relapses during adulthood [6, 7, 282‐286]. Risk factors of continued active disease during adulthood are earlier onset of NS [129, 282, 285], early relapse after onset [6, 287], FRNS or SDNS [6, 7, 284‐287], and duration of remission < 6 years [283, 288]. Accordingly, some adolescents are still using maintenance immunosuppressive therapy [285, 289] (Supplementary Table S10). Many also have experienced comorbidity from the treatment or the disease, such as hypertension, short stature, obesity, osteoporosis, cataract, dyslipidemia, infertility, and even psychiatric illness and thrombosis [6, 285, 287, 289‐292]. These conditions need to be cared for without interruption, necessitating appropriate transition when the patient becomes an adult. Since a long time may be required for patients and their caregivers to prepare for transition to adult care, plans for transition should be started when the patient becomes an adolescent.

Transition is defined as a “process that involves planned efforts to prepare the patient from caregiver-directed care to self-disease management in the adult unit” according to the consensus statement on transition endorsed by ISN and IPNA [293]. For a successful transition, a young adult should be competent in self-disease management, which can be evaluated by questionnaires such as the Ready Steady Go and the Transition scale. Examples are provided in Supplementary Tables S13 and S14. Risk of nonadherence at the time of transfer from pediatric to adult care is high [294, 295] which can be aggravated if treatment policy of adult care is different from that of pediatric care. Because disease definitions, treatment protocols, and monitoring and follow-up differ between adults and children [296‐298] (Supplementary Table S15), the patient should be educated and made aware of these differences during the period of transition to ensure adaptation and adherence to adult care.

Upon transition, a decision should be made about whether to transfer the patient to a primary care physician, local adult nephrology practice, or an academic hospital center, based on the condition and history of the patient. If the patient is prepared for transition, in remission for a long period without any immunosuppressive therapy, without additional support of other members of the multidisciplinary team (psychologist, social workers, educators), and his/her kidney function and blood pressure are normal, he/she can be referred to primary care with instructions about management, health-care checks, and when to consult hospital physicians. Otherwise, the patient should be prepared for transition to adult nephrology care. Patients who require low-complexity care can be transitioned to a nephrologist in a regional center, when the treatment plan is defined and the clinical condition of the patient is stable. When in doubt, we suggest that patients be transitioned to a nephrologist in an academic center, who can decide to share management with his/her colleague in a regional center.

For uninterrupted care, the adult nephrologist needs to know the patient thoroughly by comprehensive history-taking and evaluation (Table 6).

There are few data regarding transition care focusing on patients with SSNS [299]. Considering that quite a number of patients with childhood-onset NS persistently relapse during adulthood, a formal supportive program of transition is required.

It is advised that the patient is seen jointly by the pediatric and adult nephrologist during one or more outpatient visits. A detailed history should be transferred, which should include various aspects of the disease history as listed in Table 6. Ideally, a specialized nurse or case manager is involved in transition. This person can be the person who is primarily the key liaison for the patient.

While children are instructed to check their urine regularly, and to increase drug dose in case of a positive test, relapse during adulthood is usually not as frequent as during childhood, and the relapse rate decreases with age. Many patients may have low-grade proteinuria, or develop short-lasting proteinuria during fever, infections, or exercise. In addition, the risk of severe morbidity caused by a relapse, such as hypovolemia or thrombo-embolic events is low in adults. Therefore, patients need to be educated to rely on their own observation of signs and symptoms such as foamy urine, edema, abdominal pain, instead of relying on dipstick tests to detect a relapse, which accompanies urinary change (foamy urine) and edema at later stage. However, dipstick evaluations are recommended in any case of clinically suspected relapse.

There should be a discussion on overall management, including how to monitor and manage relapse and how to modify maintenance immunosuppression. Although many patients will experience a relapse, tapering of immunosuppressive therapy should be tried at least every 2 years, although it remains a matter of trial and error. In addition, it is important to discuss the strategy to prevent relapses during infections or stress. Likewise, information on prevention of glucocorticoid deficiency should be available and clear.